Artificial Intelligence, Personalized Persuasion, and Climate Attitudes

Elena Pro, António Valentim

European Institute, London School of Economics and Political Science

The Puzzle

FoodCloud Bear
Local Market

The Puzzle

The Puzzle

Trulli

Is it working?

F:IMF_Imminence and INFORM risk data
Infogram

Why do these messages don’t work?

  • The issue:
    • Psychological distance and intertemporal discounting reduce climate urgency (Trope & Liberman, 2010).
    • Emotional disconnection from abstract risks (Weber, 2006).
  • The solution:
    • Tailored messages increase engagement (Goldberg & Gustafson, 2025).

Core Argument

Our Argument

AI can be a powerful tool for personalised persuasion in climate communication

Why AI?

  • AI can allow us to tailor messages to individuals (instead of broad demographic groups) on larger scale at lower cost
    • Overcomes limits of static, fact-based interventions
  • Evidence shows that LLMs can:
    • Influence deeply-held beliefs (Goel et al., 2024)
    • Counter conspiracy theories (Costello et al., 2024)
    • Adapt arguments to individual values

Hypotheses

H1: When compared with the control condition (information provision), interacting with the LLM will have a positive effect on all outcomes of interest: a) climate concern, b) policy support c) pro-environmental behavior, d) conversation spillover, e) information-seeking behavior, and f) subjective climate beliefs.

H2: The personalized LLM interaction will be more persuasive/effective than the non-personalized LLM interaction and of the basic information provision.

Study Design

  • Survey experiment (UK, N = 1500–2000)
  • Four treatment arms:
    1. Static/generic information provision
    2. Non-personalised AI interaction
    3. Personalised AI on unrelated topic
    4. Personalised AI on climate change

The Interaction Flow

AI diagram

An Example

Outcomes & Analysis

  • Pre/post comparison of attitudes
  • Outcomes: Concern, behaviour, policy support - happy to talk more about them if needed :)
  • Explore heterogeneity by demographics
  • Robustness: 90% & 95% CIs;
  • Follow up survey to assess long-term effects

Behavioural Outcomes: Information-Seeking

Thank you for completing the survey!

Would you like to learn more about climate change and how to take action?

Behavioural Outcomes: Political Spill-over

Take Political Action

Would you like to write a message to your local MP about climate change?

Broader Impact

A powerful, cost-effective tool for climate campaigns:

  • Climate organisations (e.g. Greenpeace, Friends of the Earth) can use AI-personalised messaging to:
    • Enhance outreach
    • Boost fundraising
    • Mobilise volunteers
    • Strengthen policy support

For climate communicators:

  • Scientists, educators, and activists can:
    • Craft emotionally resonant messages
    • Counter misinformation
    • Make climate science more accessible

Going Forward: Field Experiment

Online Version

Greenpeace Logo Greenpeace AI 💬
AI: Hi! I'm Greenpeace AI. Ask me anything about climate change 🌍
You: What are the biggest threats to biodiversity?
AI: Deforestation, pollution, and climate change are major threats. Let me explain...

Contribution to Political Science

  • Developing a tool for personalised climate communication at scale
  • Evaluate its effectiveness compared to traditional informational strategies
  • Exploring the role of AI in shaping public opinion

Thank You

Questions?

📧 e.pro@lse.ac.uk